6.5 Hydrostatic and Pneumatic Testing of Piping and Equipment
Key Takeaways
- Under ASME B31.3, hydrostatic test pressure at every point in a metallic system must be at least 1.5 times the design pressure.
- Pneumatic test pressure under ASME B31.3 must be at least 110 percent of design pressure, and a pressure relief device is required during the test.
- Hydrostatic testing is preferred because water is nearly incompressible and stores far less energy than compressed gas at the same pressure.
- During a pneumatic test the pressure is first raised to the lesser of half the test pressure or 25 psi for a preliminary check before proceeding.
- High points must be vented while filling so trapped air is removed, and the test gauge should be sized so the test pressure falls in the middle of its range.
Why the test medium matters more than the pressure
A pressure test proves a system's integrity before it goes into service. The choice between water and gas is a stored energy decision, not a convenience decision.
Water is nearly incompressible. Pressurizing a piping system with water stores very little energy; if the system fails, the pressure collapses almost instantly and the failure is a spray, not an explosion.
Compressed gas is highly compressible. The same volume at the same pressure holds orders of magnitude more stored energy, and a failure releases all of it at once. That is why hydrostatic testing is the default and pneumatic testing is used only when water is unacceptable — for example, where the system cannot tolerate moisture, where residual water would freeze, or where the structure cannot support the weight of a water fill.
ASME B31.3 test pressures
| Test | Minimum test pressure | Additional requirement |
|---|---|---|
| Hydrostatic | Not less than 1.5 times design pressure at every point in the system | Hold at test pressure at least 10 minutes, then examine joints at a reduced pressure |
| Pneumatic | Not less than 110 percent of design pressure | A pressure relief device set no higher than the lesser of the test pressure plus 50 psi or 110 percent of the test pressure |
For the pneumatic test, B31.3 also requires a staged approach: raise the pressure gradually to the lesser of one-half the test pressure or 25 psi, hold, and make a preliminary check of all joints before continuing in steps to full test pressure.
Where the design temperature exceeds the test temperature, the required hydrostatic test pressure is adjusted upward by the ratio of allowable stresses, and where components limit the test pressure the code provides alternatives. Those adjustments are engineering decisions; the mechanic's obligation is to test at the pressure stated on the test package and not to improvise.
Preparing the system
- Read the test package. It defines the test boundary, the test pressure, the medium, the hold time, and the isolation points.
- Remove or isolate what must not see test pressure: relief valves, rupture discs, expansion joints, instruments, control valves with low-rated bodies, and pumps and vessels with lower ratings than the piping. Blind or valve them out and record it.
- Install blinds at the boundary. Blind thickness must be rated for the test pressure — a slip blind sized for a P&ID line is not automatically adequate at 1.5 times design.
- Vent every high point while filling. Trapped air is a compressed gas pocket inside what is supposed to be a hydrostatic test, and it also makes the system impossible to pressurize steadily.
- Provide low-point drains for after the test.
- Support for the weight of water. A large-diameter gas line full of water can weigh many times its normal operating load, and temporary supports or spring hanger pins may be required.
- Use calibrated gauges. Two gauges are common practice, and gauge range should be chosen so the test pressure falls roughly in the middle of the range, where a gauge is most accurate. A gauge whose full scale barely exceeds the test pressure cannot be read accurately, and one rated ten times the test pressure has no resolution.
- Check the test fluid. Chloride content is limited for austenitic stainless steel systems, because chlorides cause stress corrosion cracking.
Conducting the test
- Pressurize in stages, pausing to inspect. Never bring a system to full test pressure in one continuous ramp.
- Barricade the area and keep personnel out of the line of a potential failure, especially during pneumatic tests. Nobody stands over a blind or in front of a flange while pressure is being raised.
- Do not tighten bolts, hammer joints, or make repairs under pressure. Depressurize completely, repair, then retest.
- Hold and examine. Hold at test pressure for the required time, then reduce to the examination pressure and walk the system, checking every joint.
- Record the test pressure, hold time, gauge identification and calibration date, ambient temperature, and the result. Falling pressure on a cold night may be thermal contraction rather than a leak — record the temperature so the difference can be judged.
- Depressurize slowly, open vents before draining so the system does not pull a vacuum, and dispose of test water according to the plant's environmental procedure.
Leak testing versus pressure testing
A leak test at low pressure with soap solution or a service test checks for tightness. A pressure test at 1.5 times design proves strength. They are not interchangeable, and a system that passes a soap test at 20 psi has demonstrated nothing about its ability to hold design pressure.
A process piping system built to ASME B31.3 has a design pressure of 300 psig. What is the minimum hydrostatic test pressure?
Why is hydrostatic testing preferred over pneumatic testing whenever the system can tolerate water?
During a pneumatic test, to what pressure should the system first be raised for a preliminary check of the joints?